报告人：郭先平（中山大学）

时间：2021年5月28日09:00开始

腾讯会议ID：171 808 472

会议链接：https://meeting.tencent.com/s/rkMRRiZAYvmd

摘要:This talk is on the risk-sensitive continuous-time Markov decision processes with unbounded transition and discounted costs. Different from the case of bounded transition/cost rates, the optimality equation (OE) no longer has a solution satisfying the uniform convergent condition introduced in the existing literature. Thus, we first replace the uniform convergent condition of the solution with a suitable weighted-bound. Then, we find mild conditions imposed on the primitive data of the decision processes, which not only ensure the existence of a solution to the OE but also are the generalization of the bounded transition/cost rates conditions. Furthermore, using the characterization of the weighted-bound and a novel technique, from the OE we prove the existence of an optimal policy out of the class of randomized history-dependent policies. Finally, we present two examples with unbounded transition/cost rates to illustrate our results.

简介:郭先平，男，博士，博士生导师，国家杰出青年科学基金获得者，1996年于中南大学获博士学位，2002于中山大学晋升为教授，2003年入选教育部优秀青年教师资助计划，2004年入选教育部新世纪优秀人才支持计划，2010年被评为珠江学者特聘教授，现为全国概率统计学会副理事长。曾担任国际（SCI）杂志Advances in Applied Probability，Journal of Applied Probability，Science China Mathematics，Journal of Dynamics and Games，及国内期刊《中国科学：数学》、《应用数学学报》、《应用概率统计》、《运筹学学报》等杂志编委。研究兴趣为马氏决策过程、随机博弈、风险控制、排队优化等。

邀请人：周国立

欢迎广大师生积极参与！